Background: The titration of opioids to the optimal dose without side effects is an important and essential gateway in cancer pain management. However, sufficient pain control may not be achievable until the titration is complete because the change in the dose of opioids is calculated based on the opioid demand over the previous 24 hours. A recent patient-controlled analgesia (PCA) device could perform in an optimizing B.I. mode in which the background infusion rate was adjustable in accordance with the patient’s use of a bolus button. Therefore, we aimed to compare the efficacy and safety of the conventional method and the optimizing B.I. method of PCA in the titration of intravenous morphine in patients with cancer pain. Methods: This was a prospective, randomized, open-label, active-controlled study. After randomization, the previously administered opioids were switched to intravenous morphine and dose titration was initiated. In the conventional group, the dosage of continuous infusion morphine over the next day was adjusted based on the total amount of morphine administered during the previous 24 hours. In the PCA group, morphine titration was performed by increasing or decreasing the background infusion rate immediately in accordance with the frequency and interval of pressing the bolus button. The patients received pain management in this manner for at least three days and we compared the two groups by collecting their pain-related outcomes via medical records and questionnaires. Results: A total of 39 patients were enrolled, of whom 20 were assigned to the conventional group and 19 to the PCA group. The number of breakthrough pain complaints (NRS ≥ 4) by patients to the medical staff decreased in the PCA group, showing a difference from the conventional group within 24 hours (p = 0.012). The total daily dose of morphine was increased significantly at 0–24 hours in the PCA group (p < 0.001) and then slightly increased until 72 hours. The patient-reported pain reduction rate increased at 24–48 hours (p = 0.022) and was maintained until 72 hours in the PCA group. There was no statistically significant change in the conventional group during the study period for the above items. The conversion rate to oral or transdermal opioids after 72 hours was higher in the PCA group (p = 0.010). There were no uncontrolled side effects due to morphine titration in either group. Conclusions: Patients using the PCA with optimizing B.I. mode showed faster intravenous morphine titration and higher morphine consumption than the conventional method and achieved tolerable and rapid pain relief. This study revealed that early active opioid titration is necessary for patients admitted for pain control. It also suggested that the patient's actual opioid demands for pain control may be greater than the amount of opioid used through conventional titration methods. Clinical trial information: KCT0004008.